Medical researchers conducted a study designed to evaluate the relationship between blood pressure (y), age (x1), and cholesterol (x2). They proposed a linear model. A total of 126 individuals were randomly sampled, and blood pressure, age, and cholesterol were recorded for each. The regression summary appears here:
The regression equation is
Blood Pressure = 45.5 + 1.03 Age + 0.032 Cholesterol
Predictor Coef SE Coef T P
Constant 45.54 50.73 0.90 0.371
Age 1.02866 0.03920 26.24 0.000
Cholesterol 0.0323 0.4071 0.08 0.937
S = 5.30434 R-Sq = 85.0% R-Sq(adj) = 84.7%
Analysis of Variance
Source DF SS MS F P
Regression 2 19600.5 9800.2 348.32 0.000
Residual Error 123 3460.7 28.1
Total 125 23061.2
Which predictor (if any) does not appear to be linearly correlated with blood pressure? Which reason supports your answer?
▸
All of the predictors are significant because thep-value from the global utility test is essentially zero.
▸
All of the predictors are significant because the R-Sq value is quite high (85%).
▸
Cholesterol does not appear to be linearly correlated with blood pressure because itsp- value is 0.937.
▸
Age does not appear to be linearly correlated with blood pressure because itsp-value is essentially zero.